Developing a Blood Test for Onset of Alzheimer's
Blog Jul 11, 2014
A study co-authored by Proteome Sciences and King's College London identified a set of 10 proteins in the blood which can predict the onset of Alzheimer’s disease. The largest study of its kind to date, it marks a significant step towards developing a blood test for the disease.
To better understand the study, its implications and Proteome Sciences' involvement we spoke to Dr Ian Pike COO of Proteome Sciences.
AB: Can you tell me more about the study and how this built upon past biomarker research?
Ian Pike (IP): The samples from 1148 individuals for this study took many years to collect with initial results in March 2012 that identified three biomarker panels in blood each containing between 11 and 16 proteins that can discriminate between mild cognitive impairment (MCI), Alzheimer’s disease (AD) and control groups respectively. The recent study published in Alzheimer’s and Dementia identified 16 proteins strongly associated with brain shrinkage in either MCI or AD. A combination of 10 of these proteins could predict the progression from MCI to AD within a year with an accuracy of 87%. The study was designed to evaluate the performance of 26 proteins that we have previously discovered and, in some cases replicated, in smaller studies. In particular we wanted to establish the level of performance of individual proteins and define the optimum panel for predicting the rate of progression of patients with established memory complaints but who do not, at the time of testing, have Alzheimer's disease to facilitate improved patient selection for clinical trials of AD compounds.
AB: Compared to current techniques, how much earlier can Alzheimer’s be detected?
IP: The published method is a standard blood test. Once developed further, and subject to additional replication of our findings in larger patient groups and by other laboratories the test could be administered during a routine visit to a memory clinic or hospital. The only other methods being proposed for prediction of disease progression involve either a lumbar puncture to measure the level of proteins in cerebrospinal fluid, or PET imaging where a radioactive dye that can bind to the amyloid protein in the brain is injected and the levels binding in the brain can then be detected by placing the patient in a scanning device similar to an MRI machine. Typical scan times are 30 - 60 minutes. In our study we set out to predict conversion to AD within 12 months as this is a timeframe most useful for patients engaging in clinical trials. We do not yet know how early any biomarkers can realistically predict onset of AD and trials to determine this would need to be very large and expensive.
AB: How did Proteome Sciences become involved in the research and what role did the company play?
IP: Proteome Sciences has extensive experience in blood biomarkers working with Professor Lovestone's team at KCL since 2002 when we jointly established a public/private proteomics research facility at the Institute of Psychiatry. Our first plasma biomarkers in AD were published in Brain 2006:129: 3042-50. Several of the findings were later replicated in an independent publication (Clin Applications 2008:2 by Cutler et al.). been During the last 8 years we have continued to work alongside the academic group to analyse blood samples from patients with AD, MCI and appropriate controls to identify proteins whose levels are predictive of disease, and particularly of the likely subsequent course of disease progression. We have jointly and independently undertaken confirmation studies with alternative approaches and assembled a panel of proteins we feel have the highest predictive value. We now offer research assays for several of these panels and details of the Alzheimer’s Plasma 9-plex, Alzheimer’s CSF 16-plex and other assays can be found at www.proteomics.com
AB: With a plasma protein panel for conversion to Alzheimer’s validated, what do you see as the next steps?
IP: This panel has passed a major validation step and is being developed for the selection of patients with established mild cognitive impairment who we expect to progress to mild Alzheimer's disease within 12-18 months. The study we have just published is a first assessment of the panel performance in this setting and will require further replication studies in larger numbers of patients but this need not take a long time as all of the necessary reagents and testing platforms are available. Through the recent announcement of the UK Dementias Research Platform, Government, NHS and the Medical Research Council are coordinating access to medical records from over 2 million individuals over the age of 50 and biobanked samples that can be used for such replication studies. Part of this secondary validation will involve licensing of intellectual property around the panel owned by Proteome Sciences/King's College London, to established diagnostic test manufacturers. Licensing is managed by Proteome Sciences.
AB: A number of drug trials have failed because the drugs were administered at too late a stage, do you think this discovery will result in drugs being retested?
IP: There is certainly a discussion to be had about why so many candidate drugs have failed to show significant clinical benefit in patients with established disease. It is widely accepted that earlier treatment is more beneficial since symptomatic patients with Alzheimer’s have already suffered irreparable brain damage, particularly in the hippocampus. Once brain is lost it is very difficult to have a significant benefit in memory or the other effects of Alzheimer's disease. Using populations enriched for fast progression may have shown a greater benefit with these failed drugs but I doubt if it would be possible, or even desirable, to revisit these in new trials now, when there are more candidates already entering the clinic.